Monitoring machinery performance and alerting operators to anomalous conditions that can impact performance is an important part of operating one or a fleet of machines. Relatively simple known monitoring systems lack detailed design information that would permit them to not only monitor centrifugal compressors but also analyze performance degradation online in real-time and recommend trouble-shooting steps required to localize and mitigate the performance degradation. Moreover, current monitoring systems do not typically adjust thresholds based on compressor load or other operating conditions. Using only static thresholds permits false positive alarms. Without this calculation, only static thresholds based on constant deviation from preset values is available. Moreover, rapidly changing operational conditions or very slowly changing operational conditions may make it difficult for an operator to recognize anomalous conditions or what operational changes can be made to mitigate the anomalous conditions.